Table 2 Reduction in number of features by using proposed algorithm for different BIQA techniques on LIVE database
From: Distortion-specific feature selection algorithm for universal blind image quality assessment
BIQA technique | Total number of features | Proposed feature selection | Total processing time | |||||
|---|---|---|---|---|---|---|---|---|
| Â | Â | FF | GB | JP2KC | JPEG | WN | All features | Proposed feature selection |
BRISQUE [41] | 36 | 22 | 23 | 17 | 19 | 24 | 0.2215 | 0.2184 |
BLIINDS II [39] | 24 | 12 | 14 | 13 | 12 | 15 | 70.459 | 70.450 |
GM-LOG [30] | 40 | 22 | 20 | 24 | 24 | 25 | 0.2246 | 0.2180 |
SSEQ [45] | 12 | 5 | 5 | 4 | 6 | 6 | 2.0634 | 2.0627 |
DIIVINE [38] | 88 | 50 | 39 | 38 | 33 | 57 | 23.083 | 23.038 |
CurveletQA [31] | 12 | 6 | 6 | 6 | 6 | 10 | 2.8965 | 2.8938 |